agentleFS
pandazkipandazki/pneuma-skills/.agents/skills/showcase/SKILL.md

showcase

Skill161 starsChanged 11 days agoMITView on GitHub

Generate or refresh a Pneuma mode's launcher showcase: metadata, hero image, and feature images. Use when the user requests showcase materials.

  • Reads credentials

4 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions4 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 4

    1. BenjaminG/ai-skills★ 4same name
    2. mikec-git/showcase-skill★ 2same name
    3. Thimorrow/claude-showcase-skill★ 0same name
    4. vipink1203/vk-skills★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository8 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 3 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read the mode's manifest and skill to understand:

    line 17
  2. Distinctive:

    After generating all images:

    line 82
  3. Very specific:

    Verify showcase.json references match actual filenames

    line 84
  4. Distinctive:

    Read the images directly to understand the exact visual style before generating.

    line 96

“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

/showcase — Generate Mode Showcase Materials

Generate showcase marketing images and metadata for a Pneuma mode's launcher gallery.

Input

The user provides a mode name (e.g. remotion, draw). If not specified, ask.

Step 1: Analyze the Mode

Read the mode's manifest and skill to understand:

  • modes/<name>/manifest.ts — displayName, description, icon, key features
  • modes/<name>/skill/ — what the mode does, its workflow, unique capabilities
  • modes/<name>/showcase/showcase.json — existing showcase data (if any)

Identify 3 key selling points — the features that would make someone want to try this mode. Think from a user's perspective: what's exciting? what's unique?

Step 2: Write showcase.json

Create modes/<name>/showcase/showcase.json:

{
  "tagline": "Short punchy tagline (5-10 words)",
  "hero": "hero.png",
  "highlights": [
    {
      "title": "Feature Name",
      "description": "1-2 sentence description of the feature and why it matters.",
      "media": "feature-name.png",
      "mediaType": "image"
    }
  ]
}

Rules:

  • Tagline: short, evocative, no period at end
  • 3 highlights (occasionally 4 for complex modes)
  • Titles: 2-4 words, punchy, title case
  • Descriptions: concrete, benefit-focused, 1-2 sentences
  • Media filenames: kebab-case, descriptive

Step 3: Generate Showcase Images

Generate 4 images (1 hero + 3 highlights) using modes/_shared/scripts/generate_image.mjs through OpenRouter (OPENROUTER_API_KEY). New images default to GPT Image 2.6 Flare; pass --model sunburst when generation needs greater precision. Use edit_image.mjs for edits (Sunburst by default), or generate_image.mjs --model sunburst --image-urls ... for edits with multiple references. Inspect reference images first. If OpenRouter image generation is unavailable, complete the metadata and report which assets still need generation.

Image specs:

  • Dimensions: 1376×768 pixels (16:9)
  • Save to: modes/<name>/showcase/

Visual style — "Ethereal Tech Dark Mockup":

Each image is a conceptual UI mockup on a dark canvas that communicates one key idea at a glance. Study the reference images in existing showcase directories for the exact aesthetic:

  • Background: Deep dark (#09090b to #18181b), subtle radial gradient warmth
  • Content: Stylized UI mockups, not real screenshots — simplified, idealized representations of the mode's interface
  • Orange accent: #f97316 for interactive elements (buttons, highlights, selection states, arrows)
  • Composition patterns used across modes:
    • "Input → Output" with an orange arrow (Draw: text prompt → diagram)
    • "Before → After" split comparison (WebCraft: AI Slop vs Impeccable)
    • "Dashboard overview" with cards and stats (Evolve: proposals + metrics)
    • "Canvas with toolbar" showing the workspace (Slide: canvas + thumbnails)
    • "Feature callout" with zoomed-in UI detail (Doc: select mode highlighting)
  • Bottom key message: Some images include a one-line italic caption at the bottom summarizing the concept
  • Chat bubbles: Some images show a small chat exchange to illustrate the human-agent interaction
  • Glass/depth effects: Subtle glassmorphism on panels, rounded corners, soft shadows
  • Typography: Clean sans-serif, clear hierarchy, white text on dark

Hero image: Overview of the mode — what does the workspace look like in action? Show the most impressive/representative state.

Highlight images: Each focuses on ONE feature. The image should be self-explanatory — someone should understand the feature just by looking at it, even without reading the title.

Step 4: Verify

After generating all images:

  1. Confirm all files exist in modes/<name>/showcase/
  2. Verify showcase.json references match actual filenames
  3. Inspect the actual images for composition, readable text, dimensions, and intended visual style.
  4. Compress PNGs with the release size guidance before committing. Commit the showcase changes when the active task includes committing; otherwise leave the reviewable files in the working tree.

Reference

Study these existing showcase directories for inspiration:

  • modes/webcraft/showcase/ — design commands, anti-slop comparison, responsive preview
  • modes/draw/showcase/ — prompt→diagram, hand-drawn canvas, bidirectional selection
  • modes/slide/showcase/ — canvas with thumbnails, drag-reorder, export options
  • modes/evolve/showcase/ — evolution dashboard, analysis, proposals

Read the images directly to understand the exact visual style before generating.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.